The development of nasal vowels By George Entenman (review)

BOOK NOTICES 759 Kay model 6061b is pictured but not specifically referred to in the text) without a discussion of how to operate the machine itself. Thus the student who is interested in what the machine can do, without taking the time to go through the training to operate it himself, is in a position to cover this material in relatively few pages of reading. My own feeling is that this kind of overview is also useful for the student who is going to learn to use the sonagraph—before he gets too wrapped up in which knobs to adjust how much, and so forth. Surprisingly, Chapter 5, which deals with the interpretation of sonagrams, the central focus of the monograph, is also almost the briefest. In fact, with the exception of a few transitional remarks, the chapter consists entirely of tabular presentations of Gunnar Fant's feature system with its articulatory and acoustic characterizations, and his strategy for sonagrams whose text is unknown. Fortunately , almost one-third of the book is given over to an appendix containing excellent reproductions of analysed sonagrams; the reader can examine these, and hopefully develop some feeling for them. A step-by-step, detailed examination of a couple of sample sonagrams, with explanations of how the procedures outlined in Chapter 5 are implemented , would be very much in order. Each chapter concludes with a set of fairly prosaic questions on the material covered. Their usefulness as a control on mastery of the material is, in a classroom situation, severely reduced by a second appendix which contains the answers. References to a bibliography at the end of the book are included in the text at various points, for the student interested in pursuing particular topics in more detail. The selection in the bibliography seems somewhat strange for an introductory text: on the one hand, there is a reference to an item in a laboratory technical report—but on the other hand, obvious supplementary material which one would want to introduce to the beginning student of sonagraphy, such as Ladefoged's Elements of acoustic phonetics (1962), is not included. Apart from these criticisms, anyone who is faced with teaching a section on sonagraphy as part of a phonetics course can only envy our Austrian and German colleagues for having such a splendid little book to place in their students' hands. Frankly, I intend to rely on it heavily in developing my own classroom presentation of this material. [Gary Bevington, Northeastern Illinois University.] The development of nasal vowels. By George Entenman. (Texas linguistic forum, 7.) Austin: Department of Linguistics, University ofTexas, 1977. Pp. 156. $3.50. A common and previously uncontroversial feature of the description of the development of nasal vowels (V) in a number of languages (e.g. French, Chinese, and Portuguese) has been the claim that they derive from vowel + nasal (VN) sequences (i.e. VN-* V), first in a restricted environment and then gradually in wider environments; e.g., it has been held that the process would occur on low vowels earlier than higher vowels, or before continuants and voiceless stops earlier than voiced stops. In this monograph, Entenman terms this the 'gradualist hypothesis' and rejects it in favor of his own 'all-or-nothing' (AON) principle, which requires that a rule VN -^-V must apply, in a given language, in all environments or in none. The supposed methodological advantage of this principle is that, faced with phonetic forms such as (1) [dàt] (from an earlier form with a nasal consonant) and [¿and], one knows in advance that the underlying forms must be either (2) /dât/ /Cad/, or (3) /dant/ /¿and/, but not (4) /dät/ /¿and/. E's case for the AON principle is not very convincing. Rather than introducing new language data which unambiguously support his point, he simply shows that it is possible to re-analyze other writers' data to be in accord with AON—although this sometimes requires taking liberties with the underlying forms, reinterpreting the historical record, or ignoring counter-examples. Consider just a few of the arguments from his treatment of Hindi (47-51). Hindi possesses the words cited in 1 above (meaning 'tooth' and 'moon...